Revealing Causal Relations by Working with Control Groups
What is the method used for and for whom is it suitable?
The central and most important component of RCTs is the random allocation of observation units such as individuals or institutions to one or more treatment and control groups. The aim of random allocation is to ensure that both groups are statistically identical, i.e. individual differences between the observation units are leveled out by their random composition. The treatment group then receives a specific treatment (this can be a specific piece of equipment or a specific medication), while the control group does not. The method is suitable for determining the causal effect of a measure, for example in the evaluation of political measures.
How exactly does the method work?
After randomly dividing the observation units into treatment and control groups and carrying out the treatment trial, the effect of the treatment can be measured by comparing the treatment group with the control group. Ideally, the RCT should be designed in such a way that there is no exchange of information and no interactions between the two groups in order to avoid the possibility of spillover effects. A spillover effect is when the changed behavior (due to the intervention) in the treatment group has an effect on the control group through interactions. In addition, the so-called experimenter demand effect and the so-called Hawthorne effect are avoided by omitting information flows. These two effects describe the distorted behavior of individuals resulting from their knowledge that they are participating in an experiment or are in the treatment group. The method originates from quantitative research and is established in the natural sciences, particularly medicine. However, it is also used in the humanities.
Contact
Kontakt
